Generalised monopulse estimation and its performance

Monopulse is an established technique for phased array radars for fast and accurate angle estimation. The term monopulse stems from the fact that historically, the estimate was based on a single data snapshot. This is not essential for the method, but we keep the term for tradition. We generalise the monopulse technique as a method to determine the maximum output power of the linearly filtered data, where the filter weighting depends on the non-linearity of a set of unknown parameters. This optimum parameters are sought from a measurement of the filter output and its derivative at the parameter values close to the desired optimum. This principle is extended to the linearly filtered data sampled in time and space to estimate an arbitrary number of non-linear parameters. In particular, we consider the frequently appearing case of pre-processing (partial pre-filtering). An application of special interest is the filtering with weights determined adaptively in space and time (STAP). Finally we determine the distribution of the estimated parameters under Gaussian assumptions.